A Single-Chain Magnet with a Very High Blocking Temperature and a

Sep 14, 2015 - Syntheses, structures and magnetic properties of the lanthanide complexes of the pyrimidyl-substituted nitronyl nitroxide radical. Jian...
1 downloads 11 Views 560KB Size
Communication pubs.acs.org/IC

A Single-Chain Magnet with a Very High Blocking Temperature and a Strong Coercive Field Rafael A. Allaõ Cassaro,*,† Samira G. Reis,‡ Thamyres S. Araujo,‡ Paul M. Lahti,§ Miguel A. Novak,⊥ and Maria G. F. Vaz*,‡ †

Instituto de Química, Universidade Federal do Rio de Janeiro, 21941-909 Rio de Janeiro, Brazil Instituto de Química, Universidade Federal Fluminense, Niterói, 24020-150, Rio de Janeiro, Brazil § Department of Chemistry, University of Massachusetts, Amherst, Massachusetts 01003, United States ⊥ Instituto de Física, Universidade Federal do Rio de Janeiro, 21941-972, Rio de Janeiro, Brazil ‡

Downloaded by UNIV OF CAMBRIDGE on September 14, 2015 | http://pubs.acs.org Publication Date (Web): September 14, 2015 | doi: 10.1021/acs.inorgchem.5b01431

S Supporting Information *

SCM with a high blocking temperature and exhibits a very strong coercive field. The [M(hfac)2NaphNN]n chains were made in a single step by reacting [M(hfac)2] hydrates with NaphNN in a solvent mixture of chloroform and dry n-heptane [see the Supporting Information (SI) for details]. These compounds crystallize in the P21/n space group, with the metal ion coordinated by four oxygen atoms from two hfac ligands and two oxygen atoms from two cis-coordinated NaphNN radicals (Figure S1). The radical is bridge-coordinated to the metal ions, generating a 2-fold helical chain parallel to the crystallographic b axis (Figure 1). The

ABSTRACT: Two isostructural 1D complexes, [M(hfac)2NaphNN]n [M = MnII (1) or CoII (2); NaphNN = 1-naphthyl nitronylnitroxide], were synthesized and exhibit very strong antiferromagnetic metal−radical exchange coupling. Compound 2 shows slow magnetic relaxation behavior with a high blocking temperature (TB ≈ 13.2 K) and a very high coercive field of 49 kOe at 4.0 K.

T

he motivation to study molecular nanomagnets comes from their potential application in high-density magnetic information storage as well as the challenge of understanding their physical behavior.1−4 Among this class of compounds, one of the most promising systems is the single-chain magnet (SCM) in which the magnetic behavior arises from large uniaxial-type magnetic anisotropy associated with strong intrachain and negligible interchain magnetic interactions.5−10 However, so far, few SCMs reported in the literature display high blocking temperatures and high coercivity fields (TB > 10 K; Hc > 30 kOe), essential properties for applications.11,12 A strategy to make stable SCMs with these characteristics is very appealing, not only to investigate the physical phenomena present in such materials but also to enable the necessary steps for realistic application. Some of us recently reported a substantial change to the common strategy of making metal− radical SCM chains with alkoxyphenyl nitronylnitroxides (NNs), by using NN spin units substituted with a large polycyclic aromatic group to keep the chains structurally isolated.12,13 The [Co(hfac)2PyrNN]n chain, where PyrNN is the radical 2-(1′pyrenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazole-3-oxide1-oxyl, shows a record high blocking temperature of 14 K and a very high coercive field.12 However, this chain displayed some structural instability, attributed to the loss of solvent entrained in the lattice. Since stability is a crucial point to be considered, as stated before, we sought modification in this strategy to increase the stability while preserving the desired SCM-type behavior and high blocking temperatures. In this paper, we report two new isostructural magnetic chains, [M(hfac)2NaphNN]n, where M = MnII (1) or CoII (2) and NaphNN is 2-(1′-naphthalenyl)-4,4,5,5-tetramethyl-4,5-dihydro-1H-imidazole-3-oxide-1-oxyl. Compound 2 behaves as a © XXXX American Chemical Society

Figure 1. View along the a axis of 2. Methyl groups, hydrogen, and fluorine atoms were omitted for clarity. Atoms: carbon (black), nitrogen (blue), oxygen (red), and cobalt(II) (cyan).

metal(II) ions have distorted octahedral environments, with M− O bond lengths ranging from 2.088 to 2.178 Å and from 2.039(2) to 2.076(2) Å for 1 and 2, respectively. The O−M−O bond angles of the cis-coordinated radical oxygen atoms are 85.33(15)° for 1 and 82.47(9)° for 2. Selected additional bond lengths and angles are given in Table S2. The intrachain distances between metal ions are 7.708(2) Å for 1 and 7.628(2) Å for 2, while the shortest interchain distances between these ions are 9.721(2) and 9.801(2) Å for 1 and 2, respectively. The intrachain distances for both compounds are close to those in the [M(hfac)2(PyrNN)]n chains,12,13 but the interchain distances between metal ions are shorter by about 1.3 Å for 1 and 2, as expected from the smaller size of the naphthyl substituent as well as the absence of lattice solvent molecules. Notably, the Received: June 25, 2015

A

DOI: 10.1021/acs.inorgchem.5b01431 Inorg. Chem. XXXX, XXX, XXX−XXX

Communication

Downloaded by UNIV OF CAMBRIDGE on September 14, 2015 | http://pubs.acs.org Publication Date (Web): September 14, 2015 | doi: 10.1021/acs.inorgchem.5b01431

Inorganic Chemistry

Magnetization hysteresis curve below the blocking temperature for 2 is shown in Figure 3. At 4.0 K, a large coercive field

[Co(hfac)2(NaphNN)]n chain is stable to ambient conditions (Figure S2). The χT temperature dependence of 1 (Figure S3) shows the expected behavior for manganese(II) (SMn = 5/2) and radical (SR = 1/2) Heisenberg ferrimagnetic chains. As the temperature decreases, the χT value increases to about 76 cm3 mol−1 K at 9.1 K and then decreases to 30 cm3 mol−1 K at 2.6 K because of saturation effects. The maximum of χT changes to 199 cm3 mol−1 K at 6.9 K when the magnetic field is reduced to 100 Oe (Figure S4). An estimate of the metal-to-radical magnetic exchange constant (H = −JSi·Si) was obtained using Seiden’s model, considering data above 33 K (see the SI for details), giving J = −(298 ± 26) cm−1. This value is within the range found experimentally and by ab initio calculations for other MnII−NN compounds. Compound 2 presents a different behavior (Figures S5 and S6), with χT reaching a maximum much higher (750 cm3 mol−1 K for 100 Oe) at 45 K. Below this maximum, χT decreases due to saturation and blocking effects. In contrast with isotropic octahedral manganese(II) complexes, cobalt(II) ions in the same environment have a large magnetic anisotropy.14 Taking this into account, we used the branch chain model (also considering H = −JSi·Si) to fit the data for T ≥ 73 K, obtaining J = −(162 ± 13) cm−1 (see the SI for details). The magnetic correlation length (ξ) and χT (χT is proportional to ξ) increase when the temperature decreases, according to the expression χT = Ceff exp(Δξ/kBT), where Ceff is the effective Curie constant of the chain repeat units and Δξ is the energy required to create a domain wall, which is valid for 1D Ising-like or highly anisotropic Heisenberg systems.14,15 The exchange interaction within this Ising limit is related to Δξ as Δξ = 2|J|S2, where S is the ground-state spin quantum number.8 The domain wall formation energy is Δξ/kB = (379 ± 12) K for 2, which corresponds to an exchange strength of |J/kB| = (190 ± 6) K (132 ± 4 cm−1) obtained from the slope of a ln(χT) versus 1/T plot (Figure S7). There is a ∼20% discrepancy between this J value and the one obtained from the branch chain model (233 K), indicating that this system is not within the Ising limit. The zero-field-cooled (ZFC) magnetic susceptibility of 2 increases strongly just above 10.7 K to a maximum value at 15.7 K (Figure 2). An average blocking temperature was assigned midway between the minimum and maximum susceptibilities, at TB ≈ 13.2 K. The field-cooled (FC) susceptibility plateaus over 2.4−10.3 K, slightly increases to a maximum at 12.2 K, and then decreases until it converges with the ZFC curve; the FC and ZFC curves are coincident above 41 K.

Figure 3. Magnetization hysteresis at 4.0 K for 2. Note that this is a minor loop and is not fully saturated.

>49 kOe is observed, while at 8.0 K (see Figure S8), the coercive field is ∼30 kOe. The shape of the hysteresis curve suggests the existence of two contributions: one hysteretic with a smooth variation around zero field and a second paramagnetic-like contribution with no hysteresis, responsible for the steep change at low fields. The alternating-current (ac) susceptibility measurements for 2 were performed in zero direct-current (dc) external field (see Figure 4 and the SI). Both the χ′ (in-phase) and χ″ (out-of-

Figure 4. Temperature dependence of χ′ (upper) and χ″ (lower) of 2 under zero dc field. Relaxation times versus reciprocal temperature are shown in the inset, where the solid lines represent linear fits to the Arrhenius law.

phase) susceptibility components show frequency dependence consistent with SCM-type behavior. A maximum in χ″ is observed between 20 and 30 K, with a shoulder at higher temperatures (30−45 K) suggesting a broader second relaxation process. This second process can be better discriminated in the isothermal frequency dependence of the susceptibility (Figures S9 and S10). The magnetic relaxation times were obtained from

Figure 2. Temperature dependence of ZFC (open circles) and FC (filled circles) magnetic susceptibilities for 2 at 10 Oe. B

DOI: 10.1021/acs.inorgchem.5b01431 Inorg. Chem. XXXX, XXX, XXX−XXX

Communication

Downloaded by UNIV OF CAMBRIDGE on September 14, 2015 | http://pubs.acs.org Publication Date (Web): September 14, 2015 | doi: 10.1021/acs.inorgchem.5b01431

Inorganic Chemistry the maxima of χ″ versus T at fixed ac frequencies (20−30 K) and also by fitting χ″ versus ν at fixed temperatures by a sum of two Casimir and Du Pre expressions (see details in the SI). The relaxation times follow an Arrhenius thermal activation behavior τ = τ0 exp(Δτ/kBT) with an activation energy barrier of Δτ/kB = (398 ± 14) K and τ0 = (4 ± 3) × 10−12 s for the lower temperature maxima 20−30 K, while Δτ/kB = (344 ± 36) K and 2 × 10−10 < τ0 < 40 × 10−10 s for the relaxation process at higher temperatures. Despite the large uncertainty in τ0 of the latter, one may speculate that their sizes are about a few hundred times longer than the short ones. A standard Cole−Cole analysis of the relaxation data (Figure S11) shows partially overlapped semicircles, confirming two discernible magnetic relaxation processes. At 17.5 K, only the faster relaxation process is observed, with a Cole−Cole parameter α = 0.19 indicating a narrow distribution of relaxation times, while at 30.0 K, the slower relaxation process is dominant with α = 0.45, showing a broader distribution of relaxation times. The activation energy barriers associated with the two independent relaxation processes are nearly the same, but the τ0 values are quite different. Defects have important effects on the relaxation dynamics of SCMs,7,8,16−18 as shown by studies on SCMs intentionally doped with diamagnetic ions, which showed no dependence of the energy barrier but a clear decrease of τ0 with increased defect concentration. For nearly defect-free chains, an effectively infinite size regime can be reached, where the relaxation mechanism energy gaps follow the relationship Δτ = 2Δξ + ΔA, where ΔA is the anisotropy energy. In a finite size regime, the energy gap relationship is Δτ = Δξ + ΔA.8 In the case of 2, the experimentally obtained Δτ and Δξ show clearly that an effective infinite size regime is not reached (Δτ < 2Δξ). The presence of two relaxation processes in compound 2 is attributed to the coexistence of a broad distribution of longer chains with a slower relaxation process (longer τ0) with a dominant narrower distribution of shorter chains having faster relaxation (shorter τ0). In existing theories for one-dimensional relaxation models, the relationship between the exchange energy (J) and magnetic anisotropy (D) are important to determine the energy gaps Δτ, Δξ, and ΔA.7,8 In the Ising limit, the anisotropy is larger than the magnetic exchange energy (|D/J| > 4/3), while in the Heisenberg limit the exchange is much larger than the anisotropy (|J| ≫ |D|).8 The estimates of the D and J values obtained from the branch chain model give |D/J| ≈ 0.6, indicating that 2 is neither within the Ising limit nor within the Heisenberg limit. We expect that our results will foster further theoretical studies to determine these relationships. In conclusion, a SCM with a large coercive field and a high blocking temperature was described. No evidence of threedimensional magnetic ordering was found in the manganese derivative. For the isostructural cobalt chain, the blocking of the dynamics at temperatures below 13 K would prevent any ordering at lower temperatures. A notable advance here is that NaphNN presents much higher structural stability under ambient conditions than the previous [Co(hfac)2(PyrNN)]n chain.





Syntheses, thermogravimetry analysis of 2, and additional magnetic analyses for 1 and 2 (PDF) Supplementary X-ray crystallographic data in CIF format for 1 and 2 (CIF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]ff.br (M.G.F.V.). *E-mail: [email protected] (R.A.A.C.). Author Contributions

All the authors contributed equally to this work. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors are thankful for the financial support provided by FAPERJ and CNPq (Brazil) and LDRX-UFF (Brazil) for using its laboratory facilities. P.M.L. thanks CNPq (Visiting Researcher; 400808/2012-9).



REFERENCES

(1) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular Nanomagnets; Oxford University Press: Oxford, U.K., 2006. (2) Novak, M. A.; Sessoli, R. Quantum Tunneling of Magnetization QTM’94; NATO ASI Series 301; Kluwer Academic Press: Dordrecht, The Netherlands, 1995; pp 171−188. (3) Ardavan, A.; Rival, O.; Morton, J. J. L.; Blundell, S. J.; Tyryshkin, A. M.; Timco, G. A.; Winpenny, R. E. P. Phys. Rev. Lett. 2007, 98, 057201. (4) Mannini, M.; Pineider, F.; Sainctavit, P.; Danieli, C.; Otero, E.; Sciancalepore, C.; Talarico, A. M.; Arrio, M.-A.; Cornia, A.; Gatteschi, D.; Sessoli, R. Nat. Mater. 2009, 8, 194−197. (5) Caneschi, A.; Gatteschi, D.; Lalioti, N.; Sangregorio, C.; Sessoli, R.; Venturi, G.; Vindigni, A.; Rettori, A.; Pini, M. G.; Novak, M. A. Angew. Chem., Int. Ed. 2001, 40, 1760−1763. (6) Clèrac, R.; Miyasaka, H.; Yamashita, M.; Coulon, C. J. Am. Chem. Soc. 2002, 124, 12837−12844. (7) Bogani, L.; Vindigni, A.; Sessoli, R.; Gatteschi, D. J. Mater. Chem. 2008, 18, 4750−4758. (8) Coulon, C.; Miyasaka, H.; Clérac, R. Struct. Bonding (Berlin) 2006, 122, 163−206 Note: Relations were adapted to be in the H = −JSi·Sj formalism.. (9) Dhers, S.; Feltham, H. L. C.; Brooker, S. Coord. Chem. Rev. 2015, 296, 24−44. (10) Heintze, E.; El Hallak, F.; Clauß, C.; Rettori, A.; Pini, M. G.; Totti, F.; Dressel, M.; Bogani, L. Nat. Mater. 2013, 12, 202−206. (11) Ishii, N.; Okamura, Y.; Chiba, S.; Nogami, T.; Ishida, T. J. Am. Chem. Soc. 2008, 130, 24−25. (12) Vaz, M. G. F.; Cassaro, R. A. A.; Akpinar, H.; Schlueter, J. A.; Santos, S., Jr.; Lahti, P. M.; Novak, M. A. Chem. - Eur. J. 2014, 20, 5460− 5467 Note: The H = −2JSi·Sj formalism was used.. (13) Vaz, M. G. F.; Allão, R. A.; Akpinar, H.; Schlueter, J. A.; Santos, S., Jr.; Lahti, P. M.; Novak, M. A. Inorg. Chem. 2012, 51, 3138−3145. (14) Wei, H.; Wan, F.; Chen, Z. Sci. China, Ser. B: Chem. 2005, 48, 402−414 and references therein. (15) Loveluck, J. M.; Lovesey, S. W.; Aubry, S. J. Phys. C: Solid State Phys. 1975, 8, 3841−3856. (16) Nakamura, K.; Sasada, T. J. Phys. C: Solid State Phys. 1978, 11, 331−343. (17) Bogani, L.; Caneschi, A.; Fedi, M.; Gatteschi, D.; Massi, M.; Novak, M. A.; Pini, M. G.; Rettori, A.; Sessoli, R.; Vindigni, A. Phys. Rev. Lett. 2004, 92, 207204. (18) Pini, M. G.; Rettori, A.; Bogani, L.; Lascialfari, A.; Mariani, M.; Caneschi, A.; Sessoli, R. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 84, 094444.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.5b01431. C

DOI: 10.1021/acs.inorgchem.5b01431 Inorg. Chem. XXXX, XXX, XXX−XXX